Erythropoietin (EPO) is the main cytokine responsible for red blood cell production in the human marrow. Signal transduction through the EPO receptor consists of multiple limbs that promote the proliferation, differentiation, and survival of erythroid progenitors. Regulatory loops that modulate EPO signaling are not fully understood, but could prove critical in providing new therapeutic approaches to EPO-refractory anemias. One such poorly understood loop is the regulation of EPO-driven erythropoiesis by iron. Previously presented work has established that iron deprivation acts in a lineage specific manner to block erythroid development and inhibit the activity of aconitase, a Krebs cycle enzyme that interconverts citrate and isocitrate. In the current studies, the causal relationship between aconitase inhibition and erythroid blockade was characterized in normal donor-derived mobilized human primary hematopoietic CD34+ cells and in adult wild type C57BL/6 mice. We hypothesized that critical thresholds of aconitase activity are required for specific facets of EPO signaling during discrete phases of erythroid development. CD34+ cells in erythroid-promoting culture conditions (EPO and iron) were strongly growth inhibited in a dose-dependent manner when treated with fluoroacetate, a reversible aconitase inhibitor. Interestingly, there was no increase in overall cell death in the growth impaired cultures, providing the first evidence that growth and survival signals in primary erythroid precursors can be dissociated. Aconitase blockade also inhibited EPO-dependent erythroid maturation, with decreased glycophorin A upregulation, diminished CD34 downregulation, and a sharp decrease in globin chain protein levels after 5 days of culture. Biochemical analysis of known EPO targets showed no alterations in the phosphorylation status of STAT5, Akt and PKCalpha, effectively ruling out a role for the corresponding pathways in the impaired growth and differentiation. Moreover, intracellular ATP levels were unaffected by aconitase inhibition, and alterations in AMPK activation, an intracellular sensor of the ATP:AMP ratio, could not be detected. These results argue against energy starvation as the cause of the observed developmental defects. Cell cycle analysis using propidium iodide showed no evidence for phase-specific arrest, excluding standard checkpoint mechanisms. Delayed addition or washout of fluoroacetate at various time points during cultures identified the existence of aconitase-dependent and subsequent aconitase-independent phases of human erythroid development. To extend these studies to in vivo erythropoiesis, C57BL/6 mice underwent fluoroacetate (n=10) or saline (n=10) treatment with continuous infusion pumps at a drug dose of 4 mg/kg/day. In vivo blockade of aconitase over a two-week period resulted in an anemia characterized by a significant decrease in mean hemoglobin (11.6 vs. 14.7 g/dL, P<0.001), hematocrit (37.3 vs. 44%, P<0.001) and red cell number (7.49 vs. 9.32 × 1012 cells/liter, P<0.001). No thrombocytopenia or neutropenia was noted. Reticulocytes were diminished (6.6 vs. 7.9%, P<0.001), and mean serum EPO levels were 3 fold higher in the treated animals (533.2 vs. 172.3 pg/ml, P<0.001). Flow cytometric analysis of the marrow erythroid compartment consistently showed accumulation of cells at a Ter119-bright C71-intermediate stage (n=3 for each group). In summary, our data establish a new function for aconitase in EPO-driven erythropoiesis that is distinct from its metabolic role in cellular energy homeostasis. Sustained aconitase activity is required for the proliferation and maturation of erythroid precursors, but does not impact survival or specific cell cycle phases. Furthermore, this requirement for aconitase activity is stage specific and restricted to an early phase of EPO-dependent erythroid development. Taken together, these results suggest the existence of a new regulatory loop in which levels of aconitase activity modulate EPO-mediated growth and maturation. This novel checkpoint could provide new targets in the treatment of many disorders of red cell production.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author